首页> 外文会议>International Conference on Artificial Intelligence and Pattern Recognition >A Recurrent Net Method to solve the Diffusion Equation in the Extracellular space of the Brain
【24h】

A Recurrent Net Method to solve the Diffusion Equation in the Extracellular space of the Brain

机译:一种恢复净方法,以解决大脑细胞外空间中的扩散方程

获取原文

摘要

The extracellular space (ECS) between cells in the brain is a porous media and diffusion in this interstitial space can be quantified from measurements based on novel micro-techniques and accurately modeled with appropriate modifications to classical equations such as the diffusion equation that is treated here. A finite-difference formula in spherical coordinates is used to approximate the governing diffusion equation, and a corresponding energy function is constructed using this formula. A recurrent neural network is then designed to minimize this energy function. Results obtained from the neural nets show excellent performance in terms of accuracy and speed. The parallelism nature of these recurrent nets may make them easier to implement on fast parallel computers and give them the speed advantage over the traditional methods for solving this problem. The results are used to study the effect of all the involved parameters on the concentration distribution and to give recommendations for efficient drug delivery in the ECS.
机译:在脑细胞之间的细胞外空间(ECS)是在该间隙空间的多孔介质和扩散可以基于新颖微技术测量来量化,并准确地通过适当的修改,以经典的方程如扩散方程是这里处理建模。球形坐标中的有限差分公式用于近似控制扩散方程,并且使用该公式构建相应的能量功能。然后设计经常性神经网络以最小化该能量功能。从神经网络获得的结果表明了在准确性和速度方面具有出色的性能。这些经常性网的平行性质可以使它们更容易在快速并行计算机上实现,并为它们提供超越传统方法,以解决这个问题。结果用于研究所有涉及参数对浓度分布的影响,并提出ECS中有效药物递送的建议。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号